Lerner Research Institute News
Read about the latest advances from Lerner Research Institute scientists, including new findings, grant awards, innovations and collaborations.
Artificial Intelligence Methodology for Alzheimer’s Disease Drug Repurposing
Dr. Cheng’s team developed an artificial intelligence methodology to uncover molecular targets involved in neuroinflammation and identify candidate therapeutics for Alzheimer’s disease.
A Cleveland Clinic-led research team has developed an artificial intelligence (AI) methodology that may help accelerate drug discovery for Alzheimer’s disease (AD). According to findings published in Genome Research, the researchers utilized an AI methodology (termed GPSnet) to systematically uncover molecular targets and networks involving AD-associated brain immune cells, including astrocytes and microglia, and prioritized two common nasal corticosteroids as candidate drugs for AD treatment.
Currently the leading cause of dementia and sixth leading cause of death in the United States, AD is expected to impact 13.8 million Americans by 2050 if disease-modifying treatments are not established. Unfortunately, because the physiological processes of AD are not well understood, drug discovery and development for AD remain stunted.
“Increasing evidence suggests neuroinflammation plays crucial roles in AD pathogenesis, but broad anti-inflammatory therapies have not been clinically efficacious,” said lead study author Feixiong Cheng, PhD, Genomic Medicine. “These observations indicate that improved understanding of how brain immune cells affect inflammation could translate into identification of novel drug targets.”
Constructing a molecular network for AD-associated brain immune cells
The researchers specifically investigated two brain cell subtypes that have been implicated in AD: disease-associated microglia (DAM) and disease-associated astrocytes (DAA).
Using sequencing technologies, they measured gene expression levels for DAM and DAA in preclinical AD models and patient brain samples. They integrated the data into a model of the human interactome (the complex network of proteins and protein-protein interactions that influence cellular function and disease progression), which enabled them to explore the underlying disease mechanisms linking DAM and DAA.
“Our analysis of the DAM and DAA networks uncovered significant molecular relationships in AD pathogenesis and progression, including several overlapping genes as well as multiple altered immune pathways and AD-related pathways,” Dr. Cheng noted. “These findings provide insights into the intercellular communication between microglia and astrocytes and suggest that these cells may trigger neuroinflammation in AD by a specific molecular network manner.”
Network-based discovery of repurposable drugs
The researchers then harnessed the DAM and DAA networks to pinpoint existing drugs that could help treat AD. They hypothesized that those with the ability to significantly reverse dysregulated gene expression of DAM or DAA could serve as targets for future studies and clinical trials. Based on this hypothesis, 31 drugs were identified, of which many were anti-inflammatory agents.
They also prioritized fluticasone and mometasone, which are corticosteroid nasal sprays typically used to treat allergy-related nasal inflammation, as top candidates based on the network predictions as well as clinical data. To validate these findings, they assessed both drugs using an electronic health records database of 7.23 million patients and found fluticasone and mometasone usage to be significantly associated with reduced likelihood of AD.
“Mechanistically, we suspect that fluticasone and mometasone potentially reduce inflammation in microglia or astrocytes, resulting in protective effects against AD,” said Dr. Cheng. “In particular, we suspect that microglia-targeted approaches have a better chance of success in mild to moderate AD compared to other therapies that have been used to delay AD onset. Further studies are needed, but our results demonstrate that our methodology, if broadly applied, could significantly catalyze innovation in AD drug discovery.
Dr. Cheng and his team are currently applying cutting-edge AI and multimodal single-cell omics technology to better understand how microglia become toxic and develop microglia-targeted therapeutic strategies for late-onset sporadic AD.
Jielin Xu, PhD, a postdoctoral fellow in Dr. Cheng’s lab; Pengyue Zhang, PhD, assistant research professor at Indiana University School of Medicine; and Yin Huang, PhD, a postdoctoral fellow in Dr. Cheng’s lab, are co-first authors on the study. Dr. Cheng presented these findings at the 2021 NIH Alzheimer's Research Summit: Path to Precision Medicine for Treatment and Prevention in April. The study was supported in part by the National Institute of Aging, which is part of the National Institutes of Health.